CN105320705A - Retrieval method and device for similar vehicle - Google Patents

Retrieval method and device for similar vehicle Download PDF

Info

Publication number
CN105320705A
CN105320705A CN201410381664.7A CN201410381664A CN105320705A CN 105320705 A CN105320705 A CN 105320705A CN 201410381664 A CN201410381664 A CN 201410381664A CN 105320705 A CN105320705 A CN 105320705A
Authority
CN
China
Prior art keywords
image
vehicle
information
images
sample
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201410381664.7A
Other languages
Chinese (zh)
Other versions
CN105320705B (en
Inventor
段凌宇
黄章帅
李晨霞
黄铁军
高文
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Peking University
Original Assignee
Peking University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Peking University filed Critical Peking University
Priority to CN201410381664.7A priority Critical patent/CN105320705B/en
Publication of CN105320705A publication Critical patent/CN105320705A/en
Application granted granted Critical
Publication of CN105320705B publication Critical patent/CN105320705B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Landscapes

  • Image Analysis (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The invention provides a retrieval method and a retrieval device for a similar vehicle. The method comprises the following steps of establishing vehicle type template bases of different areas according to example images collected by a video monitoring device; determining the vehicle type information of a to-be-queried image according to the to-be-queried image including a vehicle and the area information of the to-be-queried image; establishing a query image set corresponding to the to-be-queried image according to the vehicle type information and the image information of a target database; obtaining the retrieval results of each example image in the query image set and all the images in the target database; determining the vehicle similar to the vehicle in the to-be-inquired image in the target database according to the retrieval results of all the example images in the query image set, wherein the vehicle type template base of each area includes multiple pieces of vehicle type information and vehicle type template bases corresponding to the multiple pieces of vehicle type information; the vehicle type template base corresponding to each piece of vehicle type information includes the set of multiple example images of a vehicle living example corresponding to the vehicle type information. By using the method, the robustness of the retrieval performance of the similar vehicle can be improved.

Description

Similar vehicle retrieval method and device
Technical Field
The invention relates to the technical field of intelligent traffic, in particular to a method and a device for searching similar vehicles.
Background
With the rapid development of economy, the continuous expansion of urban scale and the great increase of the number of vehicles in China, the traffic system in China is gradually going to be intelligent. The traffic monitoring video is an important data base of public security business, and plays an important role in social security and stability maintenance, illegal crime fighting and the like. Among them, it is a basic requirement to retrieve a target vehicle from a large number of surveillance videos.
In the prior art, most of similar vehicle retrieval methods perform vehicle search based on license plate numbers, the method greatly depends on the accuracy of license plate identification, and retrieval results are wrong due to different illumination conditions, license plate abrasion or intentional license plate shielding and the like. The other method is to directly compare the query image with each image in the database, the retrieval performance of the query image is greatly influenced by factors such as illumination conditions, shooting visual angles, vehicle shielding degrees and the like, and the accuracy and the robustness are not high.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides a method and a device for searching similar vehicles, which can improve the accuracy and the robustness of the similar vehicle searching performance.
In a first aspect, the present invention provides a method for searching similar vehicles, including:
establishing vehicle model template libraries in different areas according to sample images collected in a video monitoring device;
determining vehicle type information of an image to be inquired according to the image to be inquired comprising a vehicle and the area information of the image to be inquired;
establishing a query image set corresponding to the image to be queried according to the vehicle type information and the image information of a target database;
acquiring retrieval results of each sample image in the query image set and all images in the target database;
according to the retrieval results of all sample images in the query image set, determining similar vehicles in the target database and vehicles in the images to be queried;
wherein, the motorcycle type template storehouse of each region includes: the motorcycle type template storehouse of multiple motorcycle type, the motorcycle type template storehouse of every motorcycle type includes: a set of a plurality of sample images of the vehicle model; the sample image is: the method comprises the following steps of obtaining vehicle sample images under different lighting conditions, vehicle sample images at different shooting angles or vehicle sample images in different scenes.
Optionally, establishing a vehicle model template library in different areas according to sample images collected in the video monitoring device, including:
for each area, acquiring a plurality of vehicle images acquired by a video monitoring device in the area, taking the plurality of vehicle images as sample images, identifying license plate numbers in the sample images, and acquiring vehicle information corresponding to the license plate numbers from a database of a vehicle management mechanism according to the license plate numbers of the sample images, wherein the vehicle information comprises: vehicle type information;
generating a candidate vehicle type template library of the vehicle type information by using the vehicle information and the sample image;
and screening the candidate vehicle model template library to obtain a vehicle model template library of the vehicle model information, wherein each sample image in the vehicle model template library is unique.
Optionally, determining vehicle type information of the image to be queried according to the image to be queried including the vehicle and the area information of the image to be queried, including:
when the image to be inquired comprises the license plate number, identifying the license plate number in the image to be inquired, and determining the vehicle type information of the image to be inquired in a database of a vehicle management mechanism corresponding to the area information according to the license plate number and the area information of the image to be inquired;
or,
extracting a first sub-image including a vehicle in an image to be inquired;
searching whether a vehicle image matched with the first sub-image exists in a vehicle model template library corresponding to the area information of the image to be inquired;
if the vehicle image matched with the first sub-image exists, the vehicle type information of the vehicle image matched with the first sub-image is used as the vehicle type information of the image to be inquired;
if the vehicle images matched with the first sub-image do not exist, searching whether the vehicle images matched with the first sub-image exist in a vehicle model template library of all different areas;
the vehicle type information of a vehicle type template library to which the vehicle image matched with the first sub-image belongs is used as the vehicle type information of the image to be inquired;
or,
extracting a first sub-image including a vehicle in an image to be inquired;
searching whether a vehicle model template library matched with the first subimage exists in vehicle model template libraries of various vehicle models corresponding to the regional information of the image to be inquired;
if the vehicle type template library matched with the first subimage exists, the vehicle type information of the vehicle type template library matched with the first subimage is used as the vehicle type information of the image to be inquired;
if the vehicle type template library matched with the first subimage does not exist, searching whether the vehicle type template library matched with the first subimage exists in the vehicle type template libraries in all different areas;
and taking the vehicle type information of the vehicle type template library matched with the first subimage as the vehicle type information of the image to be inquired.
Optionally, the obtaining of the retrieval results of each sample image in the query image set and all images in the target database includes:
acquiring a feature descriptor of each sample image and acquiring a feature descriptor of each image in the target database;
acquiring the visual feature similarity of the feature descriptor of each sample image and the feature descriptor of each image in the target database, and forming triple information by the sample image, the images in the target database and the visual feature similarity;
the retrieval result comprises: triple information of all sample images; or, the retrieval result comprises: and (4) the three-element information of all sample images ordered according to the similarity of the visual features.
Optionally, the determining, according to the retrieval results of all sample images in the query image set, a similar vehicle to the vehicle in the image to be queried in the target database includes:
sorting the retrieval results of all sample images in the query image set according to the visual feature similarity, and selecting the image in the target database corresponding to the visual feature similarity larger than a preset first threshold value as a similar vehicle with the vehicle in the image to be queried;
or,
and normalizing the retrieval results of all sample images in the query image set, and taking the image in the target database corresponding to the normalized visual feature similarity greater than a preset second threshold value as a similar vehicle with the vehicle in the image to be queried.
Optionally, obtaining a feature descriptor of each sample image includes:
acquiring at least one local feature descriptor of a sample image, wherein the at least one local feature descriptor forms a set;
according to the selection mode of the local feature descriptors, selecting one or more local feature descriptors from all the local feature descriptors, wherein the selected one or more local feature descriptors form a first subset of the set;
reducing the dimension of the local feature descriptors in the first subset to obtain reduced-dimension local feature descriptors;
and converting the local feature descriptors after dimension reduction into global feature descriptors for expressing the visual features of the image according to a preset first rule.
Optionally, the images of the target database are images collected in a specific time period in a plurality of monitoring video devices in a specific area;
the time information includes: the earliest time point when the images are collected in the target database and the latest time point when the images are collected;
the illumination conditions are as follows: illumination information between the earliest time point to the latest time point.
In a second aspect, the present invention provides a similar vehicle search device, including:
the vehicle model template library establishing unit is used for establishing vehicle model template libraries in different areas according to sample images collected in the video monitoring device;
the vehicle type information acquisition unit is used for determining the vehicle type information of the image to be inquired according to the image to be inquired of the vehicle and the area information of the image to be inquired;
the query image set generating unit is used for establishing a query image set corresponding to the image to be queried according to the vehicle type information and the image information of the target database;
a retrieval result acquiring unit, configured to acquire retrieval results of each sample image in the query image set and all images in the target database;
the similar vehicle determining unit is used for determining similar vehicles in the target database and the vehicle in the image to be inquired according to the retrieval results of all sample images in the inquiry image set;
wherein, the motorcycle type template storehouse of each region includes: the motorcycle type template storehouse of multiple motorcycle type, the motorcycle type template storehouse of every motorcycle type includes: a set of a plurality of sample images of the vehicle model; the sample image is: the method comprises the following steps of obtaining vehicle sample images under different lighting conditions, vehicle sample images at different shooting angles or vehicle sample images in different scenes.
Optionally, the search result obtaining unit is specifically configured to
Acquiring a feature descriptor of each sample image and acquiring a feature descriptor of each image in the target database;
acquiring the visual feature similarity of the feature descriptor of each sample image and the feature descriptor of each image in the target database, and forming triple information by the sample image, the images in the target database and the visual feature similarity;
the retrieval result comprises: triple information of all sample images; or, the retrieval result comprises: and (4) the three-element information of all sample images ordered according to the similarity of the visual features.
Optionally, the similar vehicle determination unit, in particular for
Sorting the retrieval results of all sample images in the query image set according to the visual feature similarity, and selecting the image in the target database corresponding to the visual feature similarity larger than a first threshold value as a similar vehicle with the vehicle in the image to be queried;
or,
and normalizing the retrieval results of all sample images in the query image set, and taking the image in the target database corresponding to the visual feature similarity greater than a second threshold value after normalization as a similar vehicle with the vehicle in the image to be queried.
According to the technical scheme, the similar vehicle retrieval method and the similar vehicle retrieval device determine the vehicle type information of the image to be queried according to the image to be queried of the vehicle and the area information of the image to be queried by establishing the vehicle type template base in different areas, establish the query image set corresponding to the image to be queried according to the vehicle type information and the image information of the target database, obtain the retrieval results of each sample image in the query image set and all images in the target database, and determine the similar vehicle in the target database and the vehicle in the image to be queried according to the retrieval results of all sample images in the query image set, so that the accuracy and the robustness of similar vehicle retrieval performance can be improved better.
Drawings
Fig. 1 is a schematic flow chart of a similar vehicle retrieval method according to an embodiment of the present invention;
FIG. 2 is a flow chart illustrating a method for obtaining a compact global feature descriptor for an image according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of a similar vehicle search device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present invention without making any creative effort, shall fall within the protection scope of the present invention.
Fig. 1 is a schematic flow chart illustrating a similar vehicle retrieval method according to an embodiment of the present invention, and as shown in fig. 1, the similar vehicle retrieval method according to the embodiment is as follows.
101. And establishing vehicle model template libraries in different areas according to the sample images collected in the video monitoring device.
The vehicle model template library of each region in this embodiment includes: the motorcycle type template storehouse that multiple motorcycle type belongs to, the motorcycle type template storehouse of every motorcycle type includes: a set of a plurality of sample images of the vehicle model; the sample image may be: the vehicle sample images under different lighting conditions, the vehicle sample images at different shooting angles or the vehicle sample images in different scenes and the like. The sample images of the embodiment are all from vehicle images shot by a video monitoring device in a real scene.
It can be understood that the environment, the quality, the parameters and the like of the camera in different areas are different, and the vehicle model template base in different areas established in the embodiment can truly reflect the vehicle images acquired in the surveillance video scene, so that the vehicle management mechanism can conveniently manage the vehicle according to the vehicle model template base established in the embodiment.
In this embodiment, "different regions" are not limited to "different cities," but may be different regions, different counties, or even different towns in a city, and the definition of "different regions" should be according to requirements and actual situations, and this embodiment does not limit this.
It should be noted that the vehicle model template library includes: the method comprises the following steps of obtaining a plurality of vehicle sample images under different lighting conditions, vehicle sample images at different shooting angles and vehicle sample images in different scenes.
It should be noted that the vehicle model template library established in the foregoing step 101 may be used in any retrieval, and the vehicle model template library may not be established for the subsequent image to be queried, but the vehicle model template library established in the previous stage may be directly used, that is, only the vehicle model template library needs to be established once, and the vehicle model template library may be directly used in each retrieval without establishing a template library once for each retrieval.
102. And determining the vehicle type information of the image to be inquired according to the image to be inquired of the vehicle and the area information of the image to be inquired.
For example, the image to be queried in the embodiment may be an image including all vehicles, or may be a vehicle image corresponding to a part of an area of the vehicle to be queried, and the like, which is not limited in the embodiment. For example, the area information of the image to be queried is the beijing area, the tianjin area, etc., but the area in this embodiment is limited to the city, and the range of the area information may be expanded or reduced as needed.
The vehicle type information in the present embodiment may include the model of the vehicle. In other embodiments, the vehicle type information may further include a color of the vehicle, or other information of the vehicle, such as a size, etc., and the present embodiment does not limit the vehicle type information.
103. And establishing a query image set corresponding to the image to be queried according to the vehicle type information and the image information of the target database.
It should be noted that the image information of the target database is images acquired by a plurality of surveillance video devices in a specific area within a specific time period.
For example, the image information of the target database may include temporal information of the image, lighting conditions, and the like. For example, the target database is a database of images collected by video monitoring devices 8:00 to 11:00 in 26 am, 7 and 26 am, 2014 in the area of hai lake, including illumination information (such as sunny days) and time information (8: 00 to 11:00 in the morning).
In practical application, part or all of sample images in the vehicle model template library corresponding to the time information and the illumination information can be selected to form a query image set.
It can be understood that, in this step, the vehicle type information, the image information of the target database in the target retrieval area, and the vehicle type template library established in the target retrieval area can be used to establish the query image set in the target retrieval area.
104. And acquiring retrieval results of each sample image in the query image set and all images in the target database.
For example, a feature descriptor method may be used to obtain the visual feature similarity between each sample image in the query image set and the image in the target database, so as to obtain the search results of all sample images.
In a specific application, the search result may be triple information including a sample image, an image in the target database, and the visual feature similarity, where the visual feature similarity is the visual feature similarity between the sample image in the triple information and the image in the target database.
Preferably, the search result may be triple information of all sample images ordered according to the similarity of visual features.
In a specific application, when the sample image may be a sample image of a plurality of regions, the search result may be triple information of the sample images of all the regions obtained after the search results of the sample images of different regions are fused.
That is, the triplet information output in the search result may be arranged according to a preset rule and output, where the preset rule may be a rule set according to a user requirement.
105. And determining similar vehicles in the target database with the vehicles in the images to be inquired according to the retrieval results of all the sample images in the inquiry image set.
In specific application, the retrieval results of the acquired images of different samples can be mixed, sorted and output.
For example, the retrieval results of all sample images in the query image set may be sorted according to the size of the visual feature similarity, and the image in the target database corresponding to the visual feature similarity greater than a preset first threshold is selected as a similar vehicle to the vehicle in the image to be queried;
alternatively, the step 105 may further specifically be: and normalizing the retrieval results of all sample images in the query image set, and taking the image in the target database corresponding to the normalized visual feature similarity greater than a preset second threshold value as a similar vehicle with the vehicle in the image to be queried.
Of course, the image in the target database corresponding to the maximum value of the visual feature similarity may also be used as a similar vehicle of the vehicle in the image to be queried, and different methods may be selected according to the implementation manner of the image to be queried, which is not limited in this embodiment.
In this embodiment, the normalization of the search result may be a ratio of each visual feature similarity in the search result in each region to a maximum value among all visual feature similarities in the region.
It should be noted that, in the embodiment, the retrieval result is normalized, because the overall similarity of the visual features of the retrieval result of the query sample image is low or the overall similarity of some visual features is high, therefore, when the mixed ranking is output, the results with overall low similarity of the visual features are ranked behind the results with overall high similarity, which may result in inaccurate result output.
It should be noted that the manner of outputting the search result provided in this embodiment is only an example, and is not limited thereto, and in a specific application, other manners of outputting the search result of the multiple sample images by fusion/mixing and sorting may also be used.
Therefore, similar vehicles of the vehicle in the image to be inquired can be obtained, and the robustness of the searching performance is improved.
In a possible implementation scenario, the foregoing step 101 may specifically include the following sub-steps not shown in the figure:
s1011, aiming at each area, acquiring a plurality of vehicle images acquired by the video monitoring device in the area, taking the plurality of vehicle images as sample images, identifying license plate numbers in the sample images, and acquiring vehicle information corresponding to the license plate numbers from a database of a vehicle management mechanism according to the license plate numbers of the sample images, wherein the vehicle information comprises: and (4) vehicle type information.
For example, identifying the license plate number may be a technique currently known in the art, such as identifying the license plate number from the image to be queried using license plate recognition techniques. In this embodiment, a known technology may be used to identify the license plate number in the image to be queried, and the vehicle type information corresponding to the license plate number is obtained from a vehicle management database (i.e., a database of a vehicle management organization) according to the license plate number.
The vehicle sample image including the license plate number is mainly used for facilitating acquisition of vehicle information. At present, vehicle information can be obtained only in a license plate number mode, and the accuracy of the vehicle information is guaranteed.
For example, the database of the vehicle authority may include the following information: information on whether the vehicle is a model of a bmw X6 car, an audi Q7 car, a popular v6 car, etc., what color the vehicle is, black, white or silver, the purchase date of the vehicle owner's vehicle, etc.
And S1012, generating a candidate vehicle model template library of the vehicle model information by using the vehicle information and the sample image.
That is to say, for a vehicle image S in a surveillance video, the license plate number in the vehicle image S is identified, and the license plate number P of the vehicle image S is obtained; and acquiring vehicle information of the vehicle image S through the license plate number P, retrieving a vehicle type T of a vehicle corresponding to the license plate number P, and adding the vehicle image S and the vehicle type T of the vehicle into a candidate vehicle type template library TDS of the vehicle type T.
It should be noted that, in the present embodiment, the foregoing steps S1011 to S1012 may be repeatedly executed to acquire a plurality of vehicle sample images of the vehicle type T under a plurality of angles, different illuminations, and different scenes, and then add all the vehicle sample images to the candidate vehicle type template library TDS of the vehicle type T. In addition, after the candidate vehicle model template library of the vehicle model T is determined, a plurality of images can be acquired for the vehicle model T by adopting an image acquisition device, the images can include license plate numbers or the like, and the acquired images all belong to images in the candidate vehicle model template library of the vehicle model T.
It should be noted that, due to the allocation of the regions, the candidate vehicle type template library of the vehicle type T of each region may be different, for example, the candidate vehicle type template library of the vehicle type T of the beijing region, the candidate vehicle type template library of the vehicle type T of the tianjin region, and the candidate vehicle type template library of the vehicle type T of the nanjing region may be different. In the embodiment, respective candidate vehicle model template libraries can be established for different areas, so that the management of subsequent vehicle management mechanisms is facilitated.
S1013, screening the candidate vehicle model template library to obtain a vehicle model template library of the vehicle model information, wherein each sample image in the vehicle model template library is unique.
In practical application, the screening can be manual screening or automatic screening, and automatic screening is preferably realized, because the data (including images) in the vehicle model template library of each vehicle type information has thousands of pieces, manual screening causes time and labor waste, and repeated images can be automatically screened in a visual characteristic comparison mode.
In this embodiment, the number of images in the candidate vehicle type template library may be greater than or equal to the number of images in the vehicle type template library acquired last.
The data diversity in the vehicle model template library is guaranteed during screening, namely images under different representative conditions such as different angles, different scales, different colors, different shielding degrees, different shielding angles, different illumination and different weather conditions are contained as far as possible.
That is, in this step, the candidate vehicle model template library may be screened according to preset conditions. In the embodiment, the diversity of the vehicle model template library is ensured during screening, namely, the vehicle model template library comprises image samples under different representative conditions such as different angles, different scales, different colors, different shielding degrees, different shielding angles, different illumination, different weather conditions and the like as far as possible. The number of samples of each vehicle type is not specifically limited, and is at least 1, and on the premise of covering images under the condition of being as many as possible, the number of samples of each vehicle type is generally about 30, and the total number of samples of different vehicle types can be different.
In the embodiment, a good vehicle model template library can be established by screening the candidate vehicle model template library, images of various conditions are covered, and management of a vehicle management mechanism is facilitated.
In addition, a plurality of vehicle sample images can be obtained from a video monitoring device of a road, if the vehicle images comprise: a background area and a vehicle display area. The background area in this embodiment is an area that the user does not pay attention to when using the vehicle image. Thus, the vehicle sample image in the vehicle model template library acquired last may not include the background area of the vehicle image.
In another possible implementation scenario, the aforementioned step 102 may specifically include the following sub-steps not shown in the figure:
s1021, when the image to be inquired comprises the license plate number, identifying the license plate number in the image to be inquired, and determining the vehicle type information of the image to be inquired in a database of a vehicle management mechanism corresponding to the regional information according to the license plate number and the regional information of the image to be inquired.
The model information of the present embodiment may include information such as the model of the vehicle, and the color of the vehicle.
In another possible implementation scenario, the aforementioned step 102 may specifically include the following sub-steps not shown in the figure:
s102, 102a, extracting a first sub-image including the vehicle in the image to be inquired.
It is understood that, in this step, the image to be queried may or may not include the license plate number. If the image to be queried may include the license plate number, and the license plate number can be better identified, the aforementioned step S1021 is preferably adopted to obtain the vehicle type information of the image to be queried.
S102b, searching whether a vehicle image matched with the first sub-image exists in a vehicle model template library corresponding to the area information of the image to be inquired, if so, executing the following step S102c, otherwise, executing the following step S102 d.
S102, 102c, if the vehicle image matched with the first sub-image exists, the vehicle type information of the vehicle image matched with the first sub-image is used as the vehicle type information of the image to be inquired.
S102d, if the vehicle image matched with the first sub-image does not exist, searching whether the vehicle image matched with the first sub-image exists in a vehicle model template library of all different areas; and then the vehicle type information of the vehicle type template library which the vehicle image with the highest matching degree with the first sub-image belongs to can be used as the vehicle type information of the image to be inquired.
For example, a feature descriptor manner may be adopted to obtain a first similarity between the first sub-image and each image in the vehicle model template library, i.e. a first visual feature similarity;
comparing the first similarity with a preset third threshold value;
when the first similarity is larger than the third threshold, determining that the first sub-image is matched with the image corresponding to the first similarity larger than the third threshold;
and taking the vehicle type information of the vehicle type template library to which the vehicle image with the highest matching degree belongs as the vehicle type information of the image to be inquired.
In a third possible implementation, the aforementioned step 102 may specifically include the following sub-steps not shown in the figure:
s102 a', extracting a first sub-image including the vehicle in the image to be inquired.
S102b ', searching whether a vehicle model template library matched with the first sub-image exists in the vehicle model template libraries of various vehicle models corresponding to the area information of the image to be inquired, if so, executing a step S102c ', otherwise, executing a step S102d '.
In this embodiment, a feature descriptor manner may be adopted to obtain first similarities between the first sub-image and all sample images in the database of each vehicle type, so as to obtain a first similarity set; and performing mathematical statistical analysis on the first similarity set of all the sample images corresponding to each vehicle type database to obtain the visual feature similarity of the first sub-image and each vehicle type database.
For example, the average value of the first similarities of all the sample images corresponding to each vehicle model database is used as the visual feature similarity of the first sub-image and each vehicle model database.
Or, acquiring first similarity (namely first visual feature similarity) between the first sub-image and all sample images in each vehicle type database by adopting a feature description sub-mode; further, the maximum value of the first similarity of all sample images corresponding to each vehicle type database can be used as the visual feature similarity of the first sub-image and each vehicle type database;
or, acquiring first similarities of the first sub-images and all sample images in each vehicle type database by adopting a feature description sub-mode, and taking the minimum value of the first similarities of all sample images corresponding to each vehicle type database as the visual feature similarity of the first sub-images and each vehicle type database;
or, a feature descriptor manner is adopted to obtain first similarities of the first sub-images and all sample images in each vehicle type database, a heterogeneous sample analysis manner is adopted to remove isolated points in the first similarities of all sample images corresponding to each vehicle type database, an average value of the first similarities except the isolated points in all sample images corresponding to each vehicle type database is obtained, and the average value is used as the visual feature similarity of the first sub-images and each vehicle type database.
And taking the vehicle model template library corresponding to the maximum visual feature similarity as a vehicle model template library matched with the first sub-image.
The embodiment is merely an example, and in a specific application, the vehicle model template library matched with the first sub-image may also be searched in the database in other ways.
Before the visual feature similarity is obtained, the visual features of the first sub-image and the visual features of the sample images in all vehicle type template libraries need to be extracted respectively; in practical applications, the visual features of the first sub-image and each image in the pre-established database may be extracted by using a global feature descriptor, or the visual features of the first sub-image and each image in the pre-established database may be extracted by using a local feature descriptor.
S102 c', if the vehicle type template library matched with the first sub-image exists, the vehicle type information of the vehicle type template library of the vehicle image matched with the first sub-image is used as the vehicle type information of the image to be inquired.
S102 d', if there is no vehicle model template library matched with the first sub-image, searching whether there is a vehicle model template library matched with the first sub-image in the vehicle model template libraries of all different areas; and then the vehicle type information of the vehicle type template library matched with the first subimage can be used as the vehicle type information of the image to be inquired.
If one or more visual feature similarities are larger than another preset threshold, it can be determined that the first sub-image is matched with the vehicle model template library corresponding to the visual feature similarity larger than the other threshold. If all the visual feature similarity degrees are not larger than another preset threshold value, the vehicle model template library matched with the first sub-image is not considered to be available.
Correspondingly, the vehicle type information of the vehicle type template library with the highest matching degree (namely the largest visual feature similarity) with the first sub-image is used as the vehicle type information of the image to be inquired.
In another possible implementation scenario, the foregoing step 104 may specifically include the following sub-steps not shown in the figure:
s1041, obtaining a feature descriptor of each sample image, and obtaining a feature descriptor of each image in the target database.
For example, one possible implementation of obtaining a feature descriptor of a sample image is shown in fig. 2 below.
The global feature descriptor of the sample image may be obtained by the method shown in fig. 2, and correspondingly, the global feature descriptor of each image in the target database may also be obtained by the method shown in fig. 2.
S1042, obtaining the visual feature similarity of the feature descriptor of each sample image and the feature descriptor of each image in the target database, and forming triple information by the sample image, the images in the target database and the visual feature similarity.
The retrieval result comprises: triple information of all sample images; in practical applications, the search result may include: and carrying out fusion sequencing on the triple information in the search results of different samples to obtain triple information.
The sorting mode of the triple information output by the retrieval result can be set according to the requirements of users, can be sorting according to the visual feature similarity, and can also be sorting after fusion according to the sequence formed by the retrieval results of each sample image.
For example, after extracting the visual features of each sample image and the visual features of each image in the database, the visual feature similarity of each sample image and the visual features of all images in each vehicle model template library can be obtained by means of euclidean distance or march distance. Specifically, for example, the method shown in fig. 2 is used to extract the compact global descriptor, and the subsequent method for calculating the visual feature similarity may also use the hamming distance to calculate the visual feature similarity of the two images.
According to the retrieval method of the similar vehicles, vehicle type information of the image to be queried is determined according to the image to be queried of the vehicle and the area information of the image to be queried by establishing the vehicle type template base in different areas, the query image set corresponding to the image to be queried is established according to the vehicle type information and the image information of the target database, retrieval results of each sample image in the query image set and all images in the target database are obtained, and the similar vehicles in the target database and the vehicle in the image to be queried are determined according to the retrieval results of all sample images in the query image set, so that the robustness of similar vehicle retrieval performance can be improved.
Fig. 2 is a schematic flowchart illustrating a method for obtaining a compact global feature descriptor of an image according to an embodiment of the present invention, and as shown in fig. 2, the method for obtaining a compact global feature descriptor of an image according to the embodiment is as follows.
The image here may be the aforementioned sample image, or may be any image in the target database, and this implementation merely illustrates, by way of example, a method for obtaining a compact global feature descriptor.
201. At least one local feature descriptor of an image is acquired, the at least one local feature descriptor forming a set.
For example, the manner of obtaining at least one local feature descriptor of the image is an existing manner, and the local feature descriptor of the image may be extracted by using Scale-invariant feature transform (SIFT), fast-robustness features (SURF), histogram of feature gradients (HOG), and the like.
It should be understood that the SIFT or SURF extraction method may be an existing extraction method, and the embodiment is not described in detail. Generally, SIFT has dimensions of 128 dimensions and SURF has dimensions of 64 dimensions.
202. And according to the selection mode of the local feature descriptors, selecting one or more local feature descriptors from all the local feature descriptors, wherein the selected one or more local feature descriptors form a first subset of the set.
For example, if the total number of the local feature descriptors of the image is 1000, 300 local feature descriptors may be selected to form the first subset.
In addition, if the total number of local feature descriptors of the image is 150, 150 local feature descriptors may be grouped into the first subset.
For example, the first subset may be selected by a training method, for example, the SIFT is extracted for a plurality of matching image pairs and non-matching image pairs respectively. Wherein, the matched image pair refers to two images containing the same object or the same scene, and the non-matched image pair refers to two images containing different objects or different scenes. These matching image pairs and non-matching image pairs are training images and are not associated with the images mentioned in the present invention.
Obtaining probability distribution of different characteristics of the SIFT in correctly matched SIFT and mismatched SIFT through statistics; among other things, the different characteristics may include, for example: scale, direction, peak of gaussian difference, distance to the center of the image, etc.
Based on the probability distribution, calculating the probability of correct matching of the SIFTs when the characteristics of the SIFTs of the image to be operated in the step 202 are within a certain value range, and selecting one or more SIFTs from all SIFTs of the image to be operated in the step 202 according to the probability. And assuming that different characteristics of the SIFT are statistically independent, and the probability of SIFT correct matching is the product of the probabilities of SIFT correct matching calculated based on different characteristics and is used as a basis for selecting elements in the SIFT subset.
203. And reducing the dimension of the local feature descriptors in the first subset to obtain the reduced-dimension local feature descriptors.
For example, the dimension reduction process is as follows: reducing the dimension of the local feature descriptors in the first subset by using a dimension reduction matrix to obtain the dimension-reduced local feature descriptors;
the dimension reduction matrix is obtained after a preset image data set is trained in a dimension reduction mode. The image dataset may be a pre-acquired image not associated with the image of the present invention. The dimension reduction method may be a principal component analysis method, a linear discriminant analysis method, or the like.
Specifically, dimension reduction can be performed on the selected N SIFTs by using the dimension reduction matrix, and the dimensions of the SIFTs are reduced from 128 dimensions to 32 dimensions.
The purpose of reducing the dimensions of the local feature descriptors in the selected first subset is to reduce the dimensions of the global feature descriptors of the generated image, and further reduce the dimensions of the compact global feature descriptors of the finally generated image; furthermore, redundant information in the local feature descriptors in the selected first subset can be eliminated through dimension reduction operation, and therefore the image searching and matching performance is improved.
204. And converting the local feature descriptors after dimension reduction into global feature descriptors for expressing the visual features of the image according to a preset first rule.
For example, the reduced local feature descriptors may be converted according to a Fisher vector (FisherVector) generation rule to obtain a cumulative gradient vector set, and a first Fisher vector is constructed from the cumulative gradient vectors in the cumulative gradient vector set, and
and processing the cumulative gradient vector set according to a Fisher vector sparsity judgment rule, and generating a global feature descriptor for expressing the visual features of the image.
The Fisher vector sparsity discrimination rule described above may be a statistics-based sparsity discrimination rule, or the Fisher vector sparsity discrimination rule may be a probability-based sparsity discrimination rule. Note that the sparsity in the Fisher vector sparsity determination rule is: when most of the dimensions of the first Fisher vector do not contribute much to expressing the search-oriented discrimination power of the image, such first Fisher vector is said to be sparse.
In addition, the first Fisher vector can be an adaptive Fisher vector described in the art, where the adaptive Fisher vector is generated according to statistical characteristics of visual features of an image.
205. And performing data compression on the global feature descriptor to obtain a compact global feature descriptor of the image.
For example, according to the data compression rule, the values of each dimension in the global feature descriptor are represented by 1 bit. The data compression rules may include the following: if the value of a certain dimensionality in the global feature descriptor is a positive number, the binary value is 1; if the value of a dimension in the global feature descriptor is negative and zero, the binary value is 0.
In this embodiment, a simplest binarization (Binarizing) method may be adopted to perform data compression on the scalable global feature descriptor; for example, if the value of a certain dimension of the scalable global feature descriptor is a non-negative value, the corresponding position is set to 1, otherwise, the corresponding position is set to 0. In practical application, other methods for binarizing the real number vector may also be adopted, such as using a hash mapping function.
According to the method, the local feature descriptors with judgment power are selected according to the statistical characteristics of the local feature descriptors, the selected local feature descriptors are subjected to dimensionality reduction by adopting a principal component analysis method, the self-adaptive Fisher vectors of the image are generated by using the local feature descriptors after dimensionality reduction, the self-adaptive Fisher vectors are further compressed according to the sparsity of the Fisher vectors to further obtain the telescopic global feature descriptors, and finally the compact global feature descriptors are obtained by binarizing the global feature descriptors.
Fig. 3 is a schematic structural diagram of a similar vehicle search device according to an embodiment of the present invention, and as shown in fig. 3, the similar vehicle search device according to the embodiment includes: a vehicle model template library establishing unit 31, a vehicle model information acquiring unit 32, a query image set generating unit 33, a retrieval result acquiring unit 34 and a similar vehicle determining unit;
the vehicle model template library establishing unit 31 is used for establishing vehicle model template libraries in different areas according to sample images collected in the video monitoring device;
the vehicle type information acquiring unit 32 is configured to determine vehicle type information of an image to be queried according to the image to be queried including a vehicle and area information of the image to be queried;
the query image set generating unit 33 is configured to establish a query image set corresponding to the image to be queried according to the vehicle type information and the image information of the target database;
a retrieval result obtaining unit 34, configured to obtain retrieval results of each sample image in the query image set and all images in the target database;
a similar vehicle determining unit 35, configured to determine, according to the retrieval results of all sample images in the query image set, a similar vehicle in the target database as the vehicle in the image to be queried;
wherein, the motorcycle type template storehouse of each region includes: the motorcycle type template storehouse of multiple motorcycle type, the motorcycle type template storehouse of every motorcycle type includes: a set of a plurality of sample images of the vehicle model; the sample image is: the method comprises the following steps of obtaining vehicle sample images under different lighting conditions, vehicle sample images at different shooting angles or vehicle sample images in different scenes.
In addition, in a specific application, the foregoing device may further include a result output unit, not shown in the figure, and the result output unit may be configured to output the acquired similar vehicles from high to low according to the similarity.
In a specific application, the vehicle model template library establishing unit 31 is specifically used for
For each area, acquiring a plurality of vehicle images acquired by a video monitoring device in the area, taking the plurality of vehicle images as sample images, identifying license plate numbers in the sample images, and acquiring vehicle information corresponding to the license plate numbers from a database of a vehicle management mechanism according to the license plate numbers of the sample images, wherein the vehicle information comprises: vehicle type information;
generating a candidate vehicle type template library of the vehicle type information by using the vehicle information and the sample image;
and screening the candidate vehicle model template library to obtain a vehicle model template library of the vehicle model information, wherein each sample image in the vehicle model template library is unique.
In a possible implementation manner, the vehicle type information obtaining unit 32 may be specifically configured to, when the image to be queried includes a license plate number, identify the license plate number in the image to be queried, and determine, according to the license plate number and area information of the image to be queried, vehicle type information of the image to be queried in a database of a vehicle management mechanism corresponding to the area information;
in another possible implementation manner, the vehicle type information obtaining unit 32 may be further specifically configured to extract a first sub-image including a vehicle in the image to be queried;
searching whether a vehicle image matched with the first sub-image exists in a vehicle model template library corresponding to the area information of the image to be inquired;
if the vehicle image matched with the first sub-image exists, the vehicle type information of the vehicle image matched with the first sub-image is used as the vehicle type information of the image to be inquired;
if the vehicle images matched with the first sub-image do not exist, searching whether the vehicle images matched with the first sub-image exist in a vehicle model template library of all different areas;
the vehicle type information of a vehicle type template library to which the vehicle image matched with the first sub-image belongs is used as the vehicle type information of the image to be inquired;
in a third possible implementation manner, the vehicle type information obtaining unit 32 can be further specifically used for,
extracting a first sub-image including a vehicle in an image to be inquired;
searching whether a vehicle model template library matched with the first subimage exists in vehicle model template libraries of various vehicle models corresponding to the regional information of the image to be inquired;
if the vehicle type template library matched with the first subimage exists, the vehicle type information of the vehicle type template library matched with the first subimage is used as the vehicle type information of the image to be inquired;
if the vehicle type template library matched with the first subimage does not exist, searching whether the vehicle type template library matched with the first subimage exists in the vehicle type template libraries in all different areas;
and taking the vehicle type information of the vehicle type template library matched with the first subimage as the vehicle type information of the image to be inquired.
In addition, the aforementioned retrieval result obtaining unit 34 is specifically configured to obtain a feature descriptor of each sample image, and obtain a feature descriptor of each image in the target database;
acquiring the visual feature similarity of the feature descriptor of each sample image and the feature descriptor of each image in the target database, and forming triple information by the sample image, the images in the target database and the visual feature similarity;
for example, the search results include: triple information of all sample images; or, the retrieval result comprises: and (4) sorting the triple information according to the size of the similarity of the visual features.
In another implementation manner, the retrieval result obtaining unit 34 is further specifically used for
Acquiring at least one local feature descriptor of a sample image, wherein the at least one local feature descriptor forms a set;
according to the selection mode of the local feature descriptors, selecting one or more local feature descriptors from all the local feature descriptors, wherein the selected one or more local feature descriptors form a first subset of the set;
reducing the dimension of the local feature descriptors in the first subset to obtain reduced-dimension local feature descriptors;
and converting the local feature descriptors after dimension reduction into global feature descriptors for expressing the visual features of the image according to a preset first rule.
Optionally, the similar vehicle determination unit 35, in particular for
Sorting the retrieval results of all sample images in the query image set according to the visual feature similarity, and selecting the image in the target database corresponding to the visual feature similarity larger than a first threshold value as a similar vehicle with the vehicle in the image to be queried;
alternatively, in other embodiments, the similar vehicle determining unit 35 may be further specifically configured to normalize the retrieval results of all the sample images in the query image set, and use the image in the target database, which has the normalized visual feature similarity greater than a second threshold, as a similar vehicle to the vehicle in the image to be queried.
It should be noted that in this embodiment, the image of the target database is an image collected in a specific time period in a plurality of surveillance video devices in a specific area;
the time information includes: the earliest time point when the images are collected in the target database and the latest time point when the images are collected;
the illumination conditions are as follows: illumination information between the earliest time point to the latest time point.
The similar vehicle retrieval device of the embodiment can be used for retrieving similar vehicles and can improve the robustness of similar vehicle retrieval performance.
The similar vehicle search device of this embodiment may be used to implement the technical solution of the method embodiment shown in fig. 1, and the implementation principle and technical effect are similar, which are not described herein again.
Those of ordinary skill in the art will understand that: all or a portion of the steps of implementing the above-described method embodiments may be performed by hardware associated with program instructions. The program may be stored in a computer-readable storage medium. When executed, the program performs steps comprising the method embodiments described above; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.
Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.

Claims (10)

1. A method for retrieving similar vehicles, comprising:
establishing vehicle model template libraries in different areas according to sample images collected in a video monitoring device;
determining vehicle type information of an image to be inquired according to the image to be inquired comprising a vehicle and the area information of the image to be inquired;
establishing a query image set corresponding to the image to be queried according to the vehicle type information and the image information of a target database;
acquiring retrieval results of each sample image in the query image set and all images in the target database;
according to the retrieval results of all sample images in the query image set, determining similar vehicles in the target database and vehicles in the images to be queried;
wherein, the motorcycle type template storehouse of each region includes: the motorcycle type template storehouse of multiple motorcycle type, the motorcycle type template storehouse of every motorcycle type includes: a set of a plurality of sample images of the vehicle model; the sample image is: the method comprises the following steps of obtaining vehicle sample images under different lighting conditions, vehicle sample images at different shooting angles or vehicle sample images in different scenes.
2. The method of claim 1, wherein building a model template library of different regions according to sample images collected in a video monitoring device comprises:
for each area, acquiring a plurality of vehicle images acquired by a video monitoring device in the area, taking the plurality of vehicle images as sample images, identifying license plate numbers in the sample images, and acquiring vehicle information corresponding to the license plate numbers from a database of a vehicle management mechanism according to the license plate numbers of the sample images, wherein the vehicle information comprises: vehicle type information;
generating a candidate vehicle type template library of the vehicle type information by using the vehicle information and the sample image;
and screening the candidate vehicle model template library to obtain a vehicle model template library of the vehicle model information, wherein each sample image in the vehicle model template library is unique.
3. The method according to claim 1, wherein determining vehicle type information of the image to be inquired according to the image to be inquired including a vehicle and area information of the image to be inquired comprises:
when the image to be inquired comprises the license plate number, identifying the license plate number in the image to be inquired, and determining the vehicle type information of the image to be inquired in a database of a vehicle management mechanism corresponding to the area information according to the license plate number and the area information of the image to be inquired;
or,
extracting a first sub-image including a vehicle in an image to be inquired;
searching whether a vehicle image matched with the first sub-image exists in a vehicle model template library corresponding to the area information of the image to be inquired;
if the vehicle image matched with the first sub-image exists, the vehicle type information of the vehicle image matched with the first sub-image is used as the vehicle type information of the image to be inquired;
if the vehicle images matched with the first sub-image do not exist, searching whether the vehicle images matched with the first sub-image exist in a vehicle model template library of all different areas;
the vehicle type information of a vehicle type template library to which the vehicle image matched with the first sub-image belongs is used as the vehicle type information of the image to be inquired;
or,
extracting a first sub-image including a vehicle in an image to be inquired;
searching whether a vehicle model template library matched with the first subimage exists in vehicle model template libraries of various vehicle models corresponding to the regional information of the image to be inquired;
if the vehicle type template library matched with the first subimage exists, the vehicle type information of the vehicle type template library matched with the first subimage is used as the vehicle type information of the image to be inquired;
if the vehicle type template library matched with the first subimage does not exist, searching whether the vehicle type template library matched with the first subimage exists in the vehicle type template libraries in all different areas;
and taking the vehicle type information of the vehicle type template library matched with the first subimage as the vehicle type information of the image to be inquired.
4. The method of claim 1, wherein the obtaining the search result of each sample image in the query image set and all images in the target database comprises:
acquiring a feature descriptor of each sample image and acquiring a feature descriptor of each image in the target database;
acquiring the visual feature similarity of the feature descriptor of each sample image and the feature descriptor of each image in the target database, and forming triple information by the sample image, the images in the target database and the visual feature similarity;
the retrieval result comprises: triple information of all sample images; or, the retrieval result comprises: and (4) the three groups of information of all the areas are sorted according to the similarity of the visual features.
5. The method according to claim 4, wherein the determining similar vehicles in the target database to the vehicle in the image to be queried according to the retrieval results of all sample images in the query image set comprises:
sorting the retrieval results of all sample images in the query image set according to the visual feature similarity, and selecting the image in the target database corresponding to the visual feature similarity larger than a preset first threshold value as a similar vehicle with the vehicle in the image to be queried;
or,
and normalizing the retrieval results of all sample images in the query image set, and taking the image in the target database corresponding to the normalized visual feature similarity greater than a preset second threshold value as a similar vehicle with the vehicle in the image to be queried.
6. The method of claim 4, wherein obtaining the feature descriptor of each sample image comprises:
acquiring at least one local feature descriptor of a sample image, wherein the at least one local feature descriptor forms a set;
according to the selection mode of the local feature descriptors, selecting one or more local feature descriptors from all the local feature descriptors, wherein the selected one or more local feature descriptors form a first subset of the set;
reducing the dimension of the local feature descriptors in the first subset to obtain reduced-dimension local feature descriptors;
and converting the local feature descriptors after dimension reduction into global feature descriptors for expressing the visual features of the image according to a preset first rule.
7. The method according to claim 1, wherein the images of the target database are images collected in a plurality of surveillance video devices in a specific area within a specific time period;
the time information includes: the earliest time point when the images are collected in the target database and the latest time point when the images are collected;
the illumination conditions are as follows: illumination information between the earliest time point to the latest time point.
8. A similar vehicle search device, comprising:
the vehicle model template library establishing unit is used for establishing vehicle model template libraries in different areas according to sample images collected in the video monitoring device;
the vehicle type information acquisition unit is used for determining the vehicle type information of the image to be inquired according to the image to be inquired of the vehicle and the area information of the image to be inquired;
the query image set generating unit is used for establishing a query image set corresponding to the image to be queried according to the vehicle type information and the image information of the target database;
a retrieval result acquiring unit, configured to acquire retrieval results of each sample image in the query image set and all images in the target database;
the similar vehicle determining unit is used for determining similar vehicles in the target database and the vehicle in the image to be inquired according to the retrieval results of all sample images in the inquiry image set;
wherein, the motorcycle type template storehouse of each region includes: the motorcycle type template storehouse of multiple motorcycle type, the motorcycle type template storehouse of every motorcycle type includes: a set of a plurality of sample images of the vehicle model; the sample image is: the method comprises the following steps of obtaining vehicle sample images under different lighting conditions, vehicle sample images at different shooting angles or vehicle sample images in different scenes.
9. The apparatus according to claim 8, wherein the search result obtaining unit is specifically configured to obtain the search result
Acquiring a feature descriptor of each sample image and acquiring a feature descriptor of each image in the target database;
acquiring the visual feature similarity of the feature descriptor of each sample image and the feature descriptor of each image in the target database, and forming triple information by the sample image, the images in the target database and the visual feature similarity;
the retrieval result comprises: triple information of all sample images; or, the retrieval result comprises: and (4) the three groups of information of all the areas are sorted according to the similarity of the visual features.
10. Device according to claim 8, characterized in that the similar vehicle determination unit is specifically adapted to
Sorting the retrieval results of all sample images in the query image set according to the visual feature similarity, and selecting the image in the target database corresponding to the visual feature similarity larger than a first threshold value as a similar vehicle with the vehicle in the image to be queried;
or,
and normalizing the retrieval results of all sample images in the query image set, and taking the image in the target database corresponding to the visual feature similarity greater than a second threshold value after normalization as a similar vehicle with the vehicle in the image to be queried.
CN201410381664.7A 2014-08-05 2014-08-05 The search method and device of similar vehicle Active CN105320705B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410381664.7A CN105320705B (en) 2014-08-05 2014-08-05 The search method and device of similar vehicle

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410381664.7A CN105320705B (en) 2014-08-05 2014-08-05 The search method and device of similar vehicle

Publications (2)

Publication Number Publication Date
CN105320705A true CN105320705A (en) 2016-02-10
CN105320705B CN105320705B (en) 2018-09-14

Family

ID=55248105

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410381664.7A Active CN105320705B (en) 2014-08-05 2014-08-05 The search method and device of similar vehicle

Country Status (1)

Country Link
CN (1) CN105320705B (en)

Cited By (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106128090A (en) * 2016-06-30 2016-11-16 北京小米移动软件有限公司 Identify the method and device of information of vehicles
CN106897677A (en) * 2017-02-06 2017-06-27 桂林电子科技大学 A kind of vehicle characteristics classification and retrieval system and method
CN107563946A (en) * 2017-08-25 2018-01-09 厦门烽火互动科技有限公司 Bicycle parking system and its bicycle access method of application
CN107563629A (en) * 2017-08-25 2018-01-09 厦门烽火互动科技有限公司 bicycle access method
CN107610314A (en) * 2017-08-23 2018-01-19 宁波术有电子科技有限公司 A kind of method that stencil matching is accelerated in identification to crown word number
CN107944352A (en) * 2017-11-09 2018-04-20 石数字技术成都有限公司 A kind of Vehicle License Plate Recognition System and method based on picture retrieval method
CN108415937A (en) * 2018-01-24 2018-08-17 博云视觉(北京)科技有限公司 A kind of method and apparatus of image retrieval
CN108885636A (en) * 2016-03-29 2018-11-23 松下知识产权经营株式会社 Vehicle type recognition device, model recognition system and model recognizing method
CN109508395A (en) * 2018-09-26 2019-03-22 安徽四创电子股份有限公司 A method of the intelligence based on vehicle retrieval is combined related cases
CN109960742A (en) * 2019-02-18 2019-07-02 苏州科达科技股份有限公司 The searching method and device of local message
CN110019910A (en) * 2017-12-29 2019-07-16 上海全土豆文化传播有限公司 Image search method and device
CN110475503A (en) * 2017-03-30 2019-11-19 富士胶片株式会社 The working method of medical image processing device and endoscopic system and medical image processing device
CN111160183A (en) * 2019-12-20 2020-05-15 华瑞新智科技(北京)有限公司 Method and device for detecting red light running of vehicle
CN111311923A (en) * 2020-03-05 2020-06-19 浙江长元科技有限公司 Fire fighting access supervision method and system
CN112700654A (en) * 2020-12-21 2021-04-23 上海眼控科技股份有限公司 Video processing method and device, electronic equipment and storage medium
CN115352312A (en) * 2022-10-20 2022-11-18 威胜信息技术股份有限公司 Charging control method and system based on charging pile

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080292211A1 (en) * 2002-04-02 2008-11-27 Frantz Robert H Vehicle exterior examination and search system
US20090185717A1 (en) * 2008-01-21 2009-07-23 Denso Corporation Object detection system with improved object detection accuracy
CN101630361A (en) * 2008-12-30 2010-01-20 北京邮电大学 Plate number, body color and mark identification-based equipment and plate number, body color and mark identification-based method for identifying fake plate vehicles
CN103455791A (en) * 2013-08-15 2013-12-18 西安理工大学 Method for recognizing vehicle types based on linear array CCD automobile chassis imaging
CN103903005A (en) * 2012-12-25 2014-07-02 财团法人交大思源基金会 License plate image identification system and method

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080292211A1 (en) * 2002-04-02 2008-11-27 Frantz Robert H Vehicle exterior examination and search system
US20090185717A1 (en) * 2008-01-21 2009-07-23 Denso Corporation Object detection system with improved object detection accuracy
CN101630361A (en) * 2008-12-30 2010-01-20 北京邮电大学 Plate number, body color and mark identification-based equipment and plate number, body color and mark identification-based method for identifying fake plate vehicles
CN103903005A (en) * 2012-12-25 2014-07-02 财团法人交大思源基金会 License plate image identification system and method
CN103455791A (en) * 2013-08-15 2013-12-18 西安理工大学 Method for recognizing vehicle types based on linear array CCD automobile chassis imaging

Cited By (22)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108885636A (en) * 2016-03-29 2018-11-23 松下知识产权经营株式会社 Vehicle type recognition device, model recognition system and model recognizing method
CN106128090A (en) * 2016-06-30 2016-11-16 北京小米移动软件有限公司 Identify the method and device of information of vehicles
CN106897677A (en) * 2017-02-06 2017-06-27 桂林电子科技大学 A kind of vehicle characteristics classification and retrieval system and method
CN106897677B (en) * 2017-02-06 2020-04-10 桂林电子科技大学 Vehicle feature classification retrieval system and method
US11412917B2 (en) 2017-03-30 2022-08-16 Fujifilm Corporation Medical image processor, endoscope system, and method of operating medical image processor
CN110475503A (en) * 2017-03-30 2019-11-19 富士胶片株式会社 The working method of medical image processing device and endoscopic system and medical image processing device
CN107610314B (en) * 2017-08-23 2019-12-06 成都术有科技有限公司 Method for accelerating template matching in crown word number recognition
CN107610314A (en) * 2017-08-23 2018-01-19 宁波术有电子科技有限公司 A kind of method that stencil matching is accelerated in identification to crown word number
CN107563946A (en) * 2017-08-25 2018-01-09 厦门烽火互动科技有限公司 Bicycle parking system and its bicycle access method of application
CN107563629A (en) * 2017-08-25 2018-01-09 厦门烽火互动科技有限公司 bicycle access method
CN107563629B (en) * 2017-08-25 2020-07-10 厦门烽火互动科技有限公司 Bicycle storing and taking method
CN107944352A (en) * 2017-11-09 2018-04-20 石数字技术成都有限公司 A kind of Vehicle License Plate Recognition System and method based on picture retrieval method
CN110019910A (en) * 2017-12-29 2019-07-16 上海全土豆文化传播有限公司 Image search method and device
CN108415937A (en) * 2018-01-24 2018-08-17 博云视觉(北京)科技有限公司 A kind of method and apparatus of image retrieval
CN109508395A (en) * 2018-09-26 2019-03-22 安徽四创电子股份有限公司 A method of the intelligence based on vehicle retrieval is combined related cases
CN109960742A (en) * 2019-02-18 2019-07-02 苏州科达科技股份有限公司 The searching method and device of local message
CN109960742B (en) * 2019-02-18 2021-11-05 苏州科达科技股份有限公司 Local information searching method and device
CN111160183A (en) * 2019-12-20 2020-05-15 华瑞新智科技(北京)有限公司 Method and device for detecting red light running of vehicle
CN111311923A (en) * 2020-03-05 2020-06-19 浙江长元科技有限公司 Fire fighting access supervision method and system
CN112700654A (en) * 2020-12-21 2021-04-23 上海眼控科技股份有限公司 Video processing method and device, electronic equipment and storage medium
CN115352312A (en) * 2022-10-20 2022-11-18 威胜信息技术股份有限公司 Charging control method and system based on charging pile
CN115352312B (en) * 2022-10-20 2023-01-17 威胜信息技术股份有限公司 Charging control method and system based on charging pile

Also Published As

Publication number Publication date
CN105320705B (en) 2018-09-14

Similar Documents

Publication Publication Date Title
CN105320705B (en) The search method and device of similar vehicle
CN105321350B (en) Fake-licensed car detection method and device
Chaudhuri et al. Multilabel remote sensing image retrieval using a semisupervised graph-theoretic method
CN107169106B (en) Video retrieval method, device, storage medium and processor
CN104520875A (en) A method and an apparatus for the extraction of descriptors from video content, preferably for search and retrieval purpose
US10719735B2 (en) Information processing method, information processing device and video surveillance system
CN104376105A (en) Feature fusing system and method for low-level visual features and text description information of images in social media
CN105320710B (en) The vehicle retrieval method and device of resisting illumination variation
Benbihi et al. Image-based place recognition on bucolic environment across seasons from semantic edge description
CN106780639B (en) Hash coding method based on significance characteristic sparse embedding and extreme learning machine
CN107229710A (en) A kind of video analysis method accorded with based on local feature description
CN106033443B (en) A kind of expanding query method and device in vehicle retrieval
Zhang et al. Automatic discrimination of text and non-text natural images
CN105320703B (en) Enquiry expanding method and device in similar vehicle retrieval
CN112084812A (en) Image processing method, image processing device, computer equipment and storage medium
CN104317946A (en) Multi-key image-based image content retrieval method
CN115115981A (en) Data processing method, device, equipment, storage medium and computer program product
Chang et al. Unsupervised video shot detection using clustering ensemble with a color global scale-invariant feature transform descriptor
CN106777350B (en) Method and device for searching pictures with pictures based on bayonet data
CN105320704B (en) Trans-regional similar vehicle search method and device
CN110674342B (en) Method and device for inquiring target image
Hu et al. STRNN: End-to-end deep learning framework for video partial copy detection
Allouche et al. Video fingerprinting: Past, present, and future
Xi et al. Beyond context: Exploring semantic similarity for tiny face detection
He et al. Ensemble of multiple descriptors for automatic image annotation

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant